Tutorial: Set up a Safex node on Vultr

Hello Community,

I have written a tutorial on how to set up one (or many) Safex node(s) using Vultr. If you register new via the link in the article, you should receive a free credit of $50 ! Enjoy.


Update: Now including:

  • Troubleshooting commands for synchronization error, screen freeze and killed process due to insufficient RAM.
  • Creation of a server snapshot and import of snapshot in a newly deployed server to host more nodes with a few clicks.

Edit 2:
If you want to run a node from a Google instance, commands are the same but:

  • When connecting to your server, you can open the ssh directly via a browser window.
  • In order to have a full node with incoming peers, add a new firewall rule and open all ports for all instances.

Edit 3: In case you have no incoming peers on your nodes at home, open ports 29390-29394 for testnet and ports 17400-17404 for mainnet (thanks, Lomax!).


Great tutorial, would be great to adapt it for a full node on mainnet

Note: bonus tutorial in there how to top up with a mycelium bitcoin wallet :slight_smile:

nice work @cryptooli


Thanks for the kind words, Dan!

That’s a great idea, just let me know how to adapt the tutorial and I am in. Unfortunately, most of us have blown their free credits on Google, Vultr, etc. already but maybe there are some people here who have credits left or are not registered yet (Google gives you $300 credit which is great for a load of nodes. They will shut down your instances if you mine with the credit).

Yeah, I decided to put that little bonus in because I used that too and it works flawless.
(Also added to the tutorial: troubleshooting when something got stuck/frozen and snapshot function to set up more nodes)


I funded some new instances and have now running 12 nodes each on testnet and mainnet ($60/m). I mainly installed those to help resolving an issue on testnet today. This issue seems to be resolved now. Although, in case some of you were willing to fund me, I would keep the 24 nodes running. You can verify yourselves if all nodes are online on testnet and mainnet. Node IPs are as following: . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Since there was a good amount of new nodes, I have canceled the majority of my instances and have now running 3 left. The testnet has now a good amount of nodes but of course new testnet and mainnet nodes are always fine to have.


Thanks man!


Legendary Oliver, thanks dude.